Molecular Docking and Drug-Likeness for the Identification

Biomedical & Pharmacology Journal, September 2018. Vol. 11(3), p. 1301-1307

Molecular Docking and Drug-Likeness for the Identification of Inhibitory Action of Acetogenins from Annona muricata as Potential Anticancer against Hypoxia Inducible Factor 1 Alpha

Supri I. Handayani1, Rahmiati1, Lisnawati Rahmadi1, Rosmalena2 and Vivitri D. Prasasty3*

1Department of Anatomical Pathology, Faculty of Medicine, University of Indonesia, Jalan Salemba Raya 6, Jakarta 10430, Indonesia. 2Department of Medical Chemistry, Faculty of Medicine, University of Indonesia, Jalan Salemba Raya 6, Jakarta 10430, Indonesia. 3Faculty of Biotechnology, Atma Jaya Catholic University of Indonesia Jalan Jenderal Sudirman 51, Jakarta 12930, Indonesia. *Corresponding author E-mail: [email protected]

http://dx.doi.org/10.13005/bpj/1492

(Received: 03 August 2018; accepted: 30 August 2018)

Hypoxia inducible factor 1 alpha (HIF-1a) regulates cell growth and differentiation which is implicated in human cancers. HIF-1a activates its cascade carcinogenesis mechanism in cancer cells. It is well-understood that signaling is initiated by HIF-1± receptor. Overexpression of HIF-1a is associated with several different human cancers, including breast cancer, lung cancer and colon cancer. Thus, HIF-1a becomes potential target of therapeutic approach in developing HIF-1a inhibitors. The aim of this research is to investigate potential inhibitors which are known as Acetogenins (AGEs) isolated from Annona muricata against HIF-1a. In order to achieve this goal, chemical structures of all compounds were retrieved from PubChem database. Molecular docking was performed by AutoDock Vina program and the resulting binding modes were analyzed with AutoDock Tools program. Among all the compounds, murihexocin A showed the best binding modes compared to other two inhibitors based on the lowest binding energies (LBE = -7.9 kcal/mol) as high as gefitinib. This was indicating that murihexocin A has favorable interaction with the essential amino acid residues at catalytic site of HIF-1a. Drug-likeness calculation of AGEs were also performed. These in silico results could be beneficial as a compound model for further studies in-vitro and in-vivo.

Keywords: HIF-1a, Annona muricata, inhibitor, acetogenins, molecular docking, drug-likeness.

Hypoxia inducible factor-1 (HIF-1) is a sequence 52 -RCGTG-32 (which R is purine) at transcription factor that controls the expression of the response elements of hipoxia to target genes gene that involved in tumorigenesis and metastases 4. The transcription process of various genes are of malignancies 1, 2. The HIF-1a level could be activated by HIF-1, including glycolytic enzymes, enhanced within breast oncogénesis, and it closely gluconeogenesis, mediating glucose transporters, related with other tumor biomarkers 3. HIF-1a growth factors, high-energy phosphate metabolism, plays an important role in binding the consensus heme metabolism, iron transport, erythropoiesis,

This is an Open Access article licensed under a Creative Commons license: Attribution 4.0 International (CC-BY).

synthesis of nitric oxide, and regulation of with HIF-1a through molecular docking methods. vasomotor. Therefore, HIF-1 possibly promotes Computational methodologies have become a the tumor cell viability in hypoxic circumstances crucial component in drug discovery program, 5, 6. which involves identification to lead optimization. Hypoxia induces tumor cell proliferation, Molecular modeling is one of the methodologies metastasis, and the rate of cell apoptosis 7. primarily used as hit identification tool when only Moreover, HIF-1 is considered as a starting point of structure of target and its active or binding site are angiogenic process in tumor cells by transcription available 17. Docking method is an energy-based activation of cancer-related gene, such as vascular scoring function which identifies the energetically endothelial growth factor (VEGF) gene 8. The level most favorable ligand conformation that binds to of HIF-1a escalates the pathological stage which the target 18. is higher in poorly differentiated lesions than in well-differentiated lesions 3. The enhanced levels of METHOD HIF-1a are strongly related with high proliferation and enhanced ER as well as VEGF expressions 9. Protein structure preparation Therefore, the high level of HIF-1a has potency in The amino acid sequence of HIF-1a associating with further massive tumors 3. (Entry PDB code : 4z1v) was retrieved from RSCB Natural bioactive compounds which Protein Database 19. The attached ligand in the are derived from plants have been used for protein structure was removed from the binding maintaining health and remedies in many years. site and saved to a new file format: pdbqt. The The phytochemical constituents in plants have been Gasteiger charges and the solvation condition a critical pipeline for the discovery of bioactive were added to the protein structure using the substances in pharmaceutical field 10. A. muricata AutoDockTool 20. or Graviola has been greatly pressumed to have Ligand structures preparation valuable natural products that play important role Ligands which are AGEs consisting of in affecting anticancer activity 11. A. muricata eight 3D structures of natural bioactive compounds leaves have been used to investigate of numerous originally belong to A. muricata and one anticancer numbers of human diseases, including cancers 10. drug for molecular docking experiments and their The highly constituents screening are most possibly conformational energy were minimized by using affected by its major bioactive components MMFF94 force field. 8 molecules of AGEs. The known as annonaceous Acetogenins (AGEs) 12. molecule structures are retrieved from PubChem Many studies reported that isolated AGEs from database (Fig. 3). The structures were scored different extracts of the plants have significant based on their physicochemical properties under antiproliferative effects against various cancer Chemicalize (ChemAxon) and Molsoft platforms 21, cell lines 10. However, some of these studies have 22. These physicochemical properties are important defined the staple mechanism of action. Recent for developing drug candidate in every stages from in-vitro studies showed inhibition action of ethyl design to pre-clinical study. acetate extract from A. muricata leaves combating Drug likeness analysis of A. muricata bioactive lung cancer cells (A549) and colon cancer cells compounds (HCT-116 and HT-29) 10, 13. The leaf extract was 3D structures of Cinchona alkaloids capable of inducing colon carcinoma and lung were analyzed using a program based on the cancer cells apoptosis by way of mitocondrial physicochemical properties, Molsoft Drug - route. This antiproliferative effect was associated Likeness. Determination of physicochemical with cell cycle involved in the G1 phase. Moreover, properties is important in the development of drug the migration and invasion of colon cancer cells candidates in all stages ranging from study design were significantly halted by the leaf extract 14-16. through pre- clinical trials 22. The aim of this research is to determine Molecular docking of HIF-1a and A. muricata the inhibition mechanism by bioactive compounds bioactive compounds of A. muricata interact with HIF-1a. To study The three dimension structure of protein the binding interactions of bioactive compounds and ligands were prepared in pdb format. Molecular Handayani et al., Biomed. & Pharmacol. J, Vol. 11(3), 1301-1307 (2018) 1303 docking simulation was run by Autodock Vina RESULTS (Vina, The Scripps Institute) 23. The AutodockTools (ADL) was utilized in minimizing energy and Bioactive compounds isolated from A. adding the partial charges of polar hydrogens of muricata , including annomuricin A, annomuricin receptor (protein). The ligands were prepared with B, annomuricin C, annomuricin E, annomutacin, flexible torsion angles and the protein was prepared murihexocin A, murihexocin B, murihexocin C, in a static (rigid) form. Furthermore, protein and and gefitinib (Figure 1) were docked into binding ligands were kept in pdbqt formats which suitable pocket of HIF-1a. for docking simulation. The affinity binding were The lowest binding energy (LBE) to the calculated as total intermolecular energies (kcal/ target protein was murihexocin A (-7.9 kcal/mol). mol) which involved hydrogen bond, Van Der The binding interactions between A. muricata Walls force, desolvation and electrostatic energies. bioactive compounds with of HIF-1a binding On the other hand, the apropriate torsion angles pocket residues were analysed as shown in of ligand is also induced as internal ligand energy. Table 1. The docking program evaluated the lowest binding From docking result, murihexocin A energy (LBE) to obtain the best binding mode. The compound interacted hydrophobically with Thr- Root-Mean-Square Deviation (RMSD) which less 183, Trp-296, Tyr-102, His-199, Gln-147, Gln-203, than 2.0 Å was scored during running docking Glu-202, Ser-184, Pro-235, Gln-239, Tyr-103, and program. Ile-281 in the binding pocket of HIF-1a.

Fig. 1. AGEs compound of A. muricata leaves: a. annomuricin A; b. annomuricin B; c. annomuricin C; d. annomuricin E; e. annomutacin; f. murihexocin A; g. murihexocin B; h. murihexocin C; i. gefitinib 1304 Handayani et al., Biomed. & Pharmacol. J, Vol. 11(3), 1301-1307 (2018)

The molecular interaction of murihexocin DISCUSSIONS A with HIF-1a is illustrated in Fig. 2. Murihexocin A interacted with HIF-1a was stabilized by hydrogen The analysis of molecular docking has bond between nitrogen atom of carboxamide group shown that the selected bioactive compounds of Ser-184 side chain with hydrogen from hydroxyl interacted at similar site as triterpene with a group of murihexocin A. different binding mode. The calculated lowest Drug likeness properties of A. muricata binding energy (LBE) values of the protein- bioactive molecules and comercial anticancer drug ligand complexes are exhibited in Table 1. LBE molecule, gefitinib were calculated using Molsoft is combined energy of the intermolecular energy Drug – Likeness program. and the free energy torsion which indicating the likeable interactions and strong binding with

Table 1. Binding interactions between A. muricata bioactive compounds with of HIF-1± binding pocket residues

Compound LBE H- Hydrophobic Interaction with HIF-1a residues (kcal/mol) bonding

Annomuricin A -6.7 3 Tyr-93, Tyr-102, Asp-104, Leu-186, Leu-188, Gln-147, His-199, Phe-207, Ile-281, Asp-201, Arg-238, Gln-239, Trp-296, His-279 Annomuricin B -7.2 2 Thr-149, Thr-183, Ser-184, Leu-186, Trp-296, Ile-281, Gln-203, Asp-201, Arg-238, His-199, Gln-239, Tyr-102, Thr-196, His-279 Annomuricin C -7.1 2 Ser-184, Asn-294, Asn-205, Trp-296, Tyr-93, Tyr-102, Arg-238, Asp-104, Asp-237, Gln-239, Tyr-103 Annomuricin E -7.3 2 Trp-296, Gln-203, Asp-201, Arg-238, Asp-237, Pro-235, Gln-239, Tyr-102, Ile-281, Leu-188, Asn-294, Thr-196, Tyr-103, Thr-196, Lys-214 Annomutacin -6.9 0 Thr-183, Ser-184, Gln-203, Trp-296, Leu-186, Arg-238, His-199, Tyr-102, His-279, Thr-196, Ile-281, Gln-147, Trp-296, Ser-184 Murihexocin A -7.9 1 Thr-183, Trp-296, Tyr-102, His-199, Gln-147, Gln-203, Glu-202, Ser-184, Pro-235, Gln-239, Tyr-103, Ile-281 Murihexocin B -6.7 0 Tyr-93, Asp104, Tyr-102, Trp-296, Leu-166, Ser-184, Arg-238, Asp-201, Gln-203, Glu-202, Tyr-93, Asp-104, Gln-239 Murihexocin C -7.6 2 Gln-147, Thr-196, His-199, Gln-203, Arg-238, His-279, Asn-294, Ser-184, Phe-207, Gln-203, Trp-296, Leu-188 Gefitinib -7.9 2 Gln-147, Thr-196, His-279, Asn-294, Ser-164, Gln-203, Arg-238, Gln-239, Tyr-102, Tyr-103, Phe-207

Fig. 2. a. Complex interaction between murihexocin A and HIF-1a in ribbon-stick form; b. Complex interaction between murihexocin A and HIF-1a in HIF-1a binding pocket Handayani et al., Biomed. & Pharmacol. J, Vol. 11(3), 1301-1307 (2018) 1305

Table 2. Drug likeness properties of A. muricata bioactive molecules and commercial anticancer drug molecule

Compound Drug Log P Molecular TPSA Stereocenter Violation of Likeness weight (A2) number** Lipinski‘s (g/mol) Rule**

Annomuricin A -0,08 5.65 646.40 28.29 0 1 Annomuricin B 0.04 5.7 628.88 65.52 0 1 Annomuricin C -0,99 4.81 648.42 112.24 0 1 Annomuricin E -0.24 4,33 646.47 28.29 0 0 Annomutacin 0.04 4.81 647.45 65.52 0 0 Murihexocin A -0,99 4,33 628.88 112.24 0 0 Murihexocin B -0.24 4.81 648.41 28.29 0 0 Murihexocin C 0.04 4,33 648.40 65.52 0 0 Gefitinib -0.24 4,33 446.15 56.07 0 0

main amino acid residues at the binding pocket bullatacin, and bullatalicin have demonstrated of the receptor. On the other hand, LBE also significant in-vivo tumor growth inhibitory performed the intermolecular energy which was activities 24-26. calculated based on the set of total energy which involved hydrophobic interaction, hydrogen bond CONCLUSION interaction, electrostatic potential and desolvation free energy. Medicinal plants play important roles in In order to find the best lead as anticancer the development of modern therapeutic agents. This agent from A. muricata bioactive compounds, study conclusively demonstrated that A. muricata we evaluated drug likeness properties of eight was a good natural source of various phytochemical AGEs compounds compared with one commercial constituents. On the basis of our results, it can anticancer drug compound, gefitinib. It was found be concluded that the annonaceous AGEs were that all bioactive compounds had one violation of powerful phytochemicals found in A. muricata, Lipinski‘s rule of five, based on molecular weight. which offers protective effect against cancer. All molecular weight was above 500 g/mol, which In agreement with the lowest binding energy, means too big as a drug. However, according to annomuricin A, annomuricin B, annomuricin C, docking result, murihexocin A was the best lead as annomuricin E, annomutacin, murihexocin A, anticancer agent and it could be used as a model murihexocin B, murihexocin C and gefitinib were for further analysis both in-vitro and in-vivo. found -6.1, -7.2, -7.1, -7.3, -6.9, -7.9, -6.7, -7.6 Approximately 133 acetogenins (AGEs) and -7.9 kcal/mol, respectively. Murihexocin A from different medicinal plants, such as Annona showed similar LBE value with gefitinib. Thus, muricata, Annona squamosa Linn., Asiminatriloba murihexocin A was selected to be the best lead as (paw paw), and Cherimolia were reported have anticancer agent in silico. These in silico results in-vitro anticancer activities against various could be beneficial as a compound model for cancer cell lines. Some AGEs such as asimin, further experimentally in-vitro and in-vivo assays asiminecin, asiminocin, and asiminacin have shown to elucidate the exact mechanism of inhibitory exceptionally high cytotoxicity for malignancies activity and to examine its potential therapeutic in three major tissues: breast, lung, and colon. effects. Moreover, in-vivo data have been documented along with the tumor cell types, animal used, route ACKNOWLEDGEMENT of administration and dosage information. Some AGEs including annonacin, desacetyluvaricin, Authors would like to thank The Scientific 1306 Handayani et al., Biomed. & Pharmacol. J, Vol. 11(3), 1301-1307 (2018)

Committee of Department of Anatomical Pathology, flavonoids and acetogenins in Graviola (Annona Faculty of Medicine, University of Indonesia for muricata) leaves confer protection against the support to accomplish this work. prostate cancer. Carcinogenesis, 36(6): 656-665 (2015). 14. Moghadamtousi, S.Z. et al. Annona muricata REFERENCES leaves induce G1 cell cycle arrest and apoptosis through mitochondria-mediated pathway in 1. Harris, A.L. Hypoxia—a key regulatory factor human HCT-116 and HT-29 colon cancer cells. in tumour growth. Nature Reviews Cancer, 2(1): Journal of ethnopharmacology, 156: 277-289 38-47 (2002). (2014). 2. Semenza, G.L. Targeting HIF-1 for cancer 15. Pieme, C.A. et al. Antiproliferative activity therapy. Nature Reviews Cancer, 3(10): 721-732 and induction of apoptosis by Annona muricata (2003). (Annonaceae) extract on human cancer cells. 3. Bos, R. et al. Levels of hypoxia-inducible BMC complementary and alternative medicine, factor-1± during breast carcinogenesis. Journal 14: 516 (2014). of the national cancer institute, 93(4): 309-314 16. Najmuddin, S.U.F.S., Romli, M.F., Hamid, M., (2001). Alitheen, N.B. & Rahman, N.M.A.N.A. Anti- 4. Semenza, G.L. Defining the role of hypoxia- cancer effect of Annona Muricata Linn Leaves inducible factor 1 in cancer biology and Crude Extract (AMCE) on breast cancer cell line. therapeutics. Oncogene, 29(5): 625-634 (2010). BMC complementary and alternative medicine, 5. Seeber, L.M. et al. The role of hypoxia inducible 16: 311 (2016). factor-1alpha in gynecological cancer. Critical 17. Arnold, K., Bordoli, L., Kopp, J. & Schwede, T. reviews in oncology/hematology, 78: 173-184 The SWISS-MODEL workspace: a web-based (2011). environment for protein structure homology 6. Seeber, L.M. et al. Necrosis related HIF-1± modelling. Bioinformatics, 22(2): 195-201 expression predicts prognosis in patients with (2006). endometrioid endometrial carcinoma. BMC 18. Kitchen, D.B., Decornez, H., Furr, J.R. & cancer, 10(1): 307-314 (2010). Bajorath, J. Docking and scoring in virtual 7. Muz, B., de la Puente, P., Azab, F. & Azab, screening for drug discovery: methods and A.K. The role of hypoxia in cancer progression, applications. Nature reviews Drug discovery, angiogenesis, metastasis, and resistance to 3(11): 935-949 (2004). therapy. Hypoxia, 3: 83-92 (2015). 19. Berman, H., Henrick, K. & Nakamura, H. 8. Aldebasi, Y.H., Rahmani, A.H., Khan, A.A. & Announcing the worldwide protein data bank. Aly, S.M. The effect of vascular endothelial Nature Structural and Molecular Biology, growth factor in the progression of bladder cancer 10(12): 980 (2003). and diabetic retinopathy. International journal of 20. Huey, R., Morris, G.M. & Forli, S. Using clinical and experimental medicine, 6(4): 239- AutoDock 4 and AutoDock Vina with 251 (2013). AutoDockTools: A Tutorial. The Scripps Research 9. Li, Q., Ma, R. & Zhang, M. CoCl2 increases the Institute Molecular Graphics Laboratory (2012). expression of hypoxic markers HIF 1±, VEGF 21. Weber, L., Vol. 5 65-66 (Royal Soc Chemistry and CXCR4 in breast cancer MCF 7 cells. Thomas Graham House, Science Park, Milton Oncology letters, 15(1): 1119-1124 (2018). Rd, Cambridge CB4 0WF, Cambs, England, 10. Moghadamtousi, S.Z. et al. Annona muricata 2008). (Annonaceae): a review of its traditional uses, 22. Arnautova, Y.A., Abagyan, R.A. & Totrov, isolated acetogenins and biological activities. M. Development of a new physics based International journal of molecular sciences, internal coordinate mechanics force field and its 16(7): 15625-15658 (2015). application to protein loop modeling. Proteins: 11. Rosdi, M.N.M., Daud, N.N.N.N.M., Zulkifli, Structure, Function, and Bioinformatics, 79(2): R.M. & Ya’akob, H. Cytotoxic effect of Annona 477-498 (2011). muricata Linn leaves extract on Capan-1 cells. 23. Trott, O. & Olson, A.J. AutoDock Vina: improving Journal of Applied Pharmaceutical Science, 5(5): the speed and accuracy of docking with a new 45-48 (2015). scoring function, efficient optimization, and 12. Rupprecht, J.K., Hui, Y.-H. & McLaughlin, J.L. multithreading. Journal of computational Annonaceous acetogenins: a review. Journal of chemistry, 31(2): 455-461 (2010). natural products, 53(2): 237-278 (1990). 24. Mangal, M., Imran Khan, M. & Mohan Agarwal, 13. Yang, C. et al. Synergistic interactions among S. Acetogenins as potential anticancer agents. Handayani et al., Biomed. & Pharmacol. J, Vol. 11(3), 1301-1307 (2018) 1307

Anti-Cancer Agents in Medicinal Chemistry 7(6): 24-35 (2015). (Formerly Current Medicinal Chemistry-Anti- 26. Yang, R.-m. et al. Anticancer effect of total Cancer Agents), 16(2): 138-159 (2016). annonaceous acetogenins on hepatocarcinoma. 25. Qayed, W.S., Aboraia, A.S., Abdel-Rahman, Chinese journal of integrative medicine, 2: 682- H.M. & Youssef, A.F. Annonaceous acetogenins 688 (2015). as a new anticancer agent. Der Pharma Chemica,